EDITORIAL COMMENT
Biomarkers for Prognostication After Acute Coronary SyndromesNew Times and Statistics*
Christopher R. deFilippi, MD ,* and
Stephen L. Seliger, MD, MS
Division of Cardiology, University of Maryland School of Medicine, Baltimore, Maryland
Division of Nephrology, University of Maryland School of Medicine, Baltimore, Maryland
* Reprint requests and correspondence: Dr. Christopher R. deFilippi, G3K63, Division of Cardiology, University of Maryland, 22 South Greene Street, Baltimore, Maryland 21201 (Email: cdefilip{at}medicine.umaryland.edu).
Key Words: acute coronary syndrome • biomarkers • risk assessment • stable coronary artery disease
Six years ago, Morrow and Braunwald (1) recognized the importance of a multimarker strategy including assessment of myocyte necrosis by cardiac troponins, hemodynamic stress by B-type natriuretic peptide (BNP) or amino terminal pro–B-type natriuretic peptide (NT-proBNP), inflammation by high-sensitivity C-reactive protein (hsCRP), and vascular damage as measured by a measure of renal function or proteinuria. In the interim, additional studies have confirmed the prognostic information and synergism provided by each arm (2). Furthermore, improvements in the sensitivity of existing markers and the addition of new markers to each arm have added incrementally to risk stratification models developed on the basis of the prototypical markers (3–6). Early models gave equal weight to each biomarker (troponin, hsCRP, and BNP) and simply tallied the number of positive markers (7). A more recent analysis from the GUSTO (Global Utilization of Strategies To Open occluded arteries) IV study found that creatinine clearance and troponin elevation provided the greatest relative contribution to risk (1-year mortality or 30-day mortality/myocardial infarction) beyond traditional risk factors and ST-segment depression as compared with NT-proBNP and hsCRP (8). Recognition of the importance of troponin levels and renal function for prognostication is incorporated into guidelines and widely used in practice to guide management (9). Does this mean that measurement of hsCRP and natriuretic peptides should be relegated to a second tier in the non–ST-segment elevation acute coronary syndromes (ACS) patients? Timing is perhaps critical. Typically, biomarker studies of ACS patients have collected and reported findings on the basis of a single sample collected early in the course of an ACS hospital stay. Although some advocate measurement of hsCRP in the first 48 h of ACS, optimal timing for risk stratification with this marker might be at least 30 days after the event, with the ability to then interpret levels after the impact of the near ubiquitous use of statin therapy (10,11). Interestingly, NT-proBNP levels also become more predictive of death the further out from the index event, indicating that measurement at the time of initial presentation, although predictive, is less than optimal (12). To put this in perspective, a panel of similar cardiac and renal risk markers, including a sensitive troponin I assay, were predictive of cardiovascular death with rigorous statistical methodology (as outlined in more detail in the following text) in community-dwelling elderly men with and without known cardiovascular disease, indicating that there is substantial prognostic synergistic information that can be derived from all of these biomarkers when patients are in a steady state some distance from their ACS event (13). However, not only is the timing of measurement of a biomarker important, but the change or absence of change in level can also carry relevant information. Unlike traditional clinical risk factors, which are in large part static and often represent decades of prior exposure, biomarker measurement is dynamic with changes over time common, and this change in level (or lack of it) can be as prognostic as the initial level itself (14,15). For example, Eggers et al. (14) have previously shown in the FRISC (FRagmin and fast revascularization during InStability in Coronary artery disease) II study patients that even a very minor persistent elevation of cardiac troponin I over 6 months carries a higher 5-year mortality than those patients with only transient elevations.
With emerging data about the importance of temporal changes in biomarker levels (which is likely to get even more complex with the introduction of increasingly sensitive troponin assays), clinicians are facing a dilemma as to when to consider measuring which marker and how often (if at all). The present study by Eggers et al. (16) in this issue of the Journal attempts to rigorously address at least this first issue in 2 ways. First, the investigators measure a panel of biomarkers at randomization, 6 weeks, and 6 months for 5-year outcomes of death, myocardial infarction, or both. Ultimately they conclude that only an NT-proBNP level at the intermediate time point of 6 weeks added significant prognostic information to statistical models containing conventional risk indicators. This conclusion is paradoxical to earlier publications in the same population and needs to be interpreted in light of the application of newer statistical tests that are being used with increasing frequency in the assessment of the incremental value of biomarker testing to conventional risk indicators (12,14). The 2 tests included in this analysis that warrant further discussion are the c-statistic and the net reclassification improvement (NRI) (17).
The c-statistic represents the area under the receiver-operating curve, a plot of sensitivity versus 1 – specificity over the full range of an individual biomarker or (when applied to a multivariate regression model) the range of predicted outcome probabilities. The c-statistic is a commonly employed measure of discrimination (i.e., the capacity of a biomarker or statistical model to successfully discriminate those who experience an outcome from those who do not). By comparing the c-statistics for 2 prediction models—1 with the "novel" biomarkers and 1 with standard or traditional risk factors—one can quantify the incremental gain in prognostic accuracy.
Although the c-statistic provides a summary of prognostic accuracy that is independent of a specific cut-point of a biomarker, recently, Cook (18) has questioned its appropriateness in identifying biomarkers of potential clinical utility. First, it is relatively insensitive for detecting moderately-sized effects that might still have clinical relevance; for example, Cook demonstrated that widely accepted, established cardiovascular risk factors such as systolic blood pressure and cholesterol are associated with small incremental gains in the c-statistic (0.03 to 0.04) for the prediction of cardiovascular events. Considered in this context, the 0.03 gain in c-statistic observed after adding NT-proBNP (at 6 weeks or 6 months after ACS) to clinical risk indicators in FRISC II might be considered clinically important. An additional limitation of the c-statistic is that it does not take into account the magnitude of change in estimated risk associated with adding a biomarker or set of biomarkers to a traditional risk profile, although it is precisely this risk change that is of relevance to the clinician. For example, a clinician is more likely to modify his or her clinical management if a patient is reclassified to a markedly higher risk category after measurement of a biomarker or set of markers.
The NRI, used by Eggers et al. (16) in this study, is an estimate of this type of reclassification—the extent to which a biomarker correctly reclassifies individuals to a higher or lower risk of a specific outcome. The NRI was designed for studies in which established clinically relevant risk categories were considered—for example, the low-, intermediate-, and high-risk categories defined by the Framingham Risk Score. With these categories, Eggers et al. (16) demonstrate that 11% of patients are correctly reclassified into higher or lower risk categories; interestingly, much of this improvement resulted from patients without an event who were previously classified as high risk (with traditional risk factors) being correctly reclassified as low or intermediate risk on the basis of their NT-proBNP levels. This 11% net reclassification compares favorably with the NRI values reported in other biomarker studies, including hsCRP in the Physicians Health Study for the prediction of cardiovascular disease events (5.4%) and troponin-I or NT-proBNP for prediction of cardiovascular disease mortality in older Swedish men (12% to 15%) (13,19). Whether 11% net reclassification justifies the routine measurement of NT-proBNP 6 weeks after an ACS event depends in part on cost-effectiveness considerations and whether a reclassification to a lower risk category has the same ramifications for clinical care as reassignment to a higher risk category.
The current study highlights the strengths and limitations of the current statistical methods for quantifying prognostic accuracy of new biomarkers and risk factors. Metrics such as the c-statistic, the NRI, and related measures of discrimination, reclassification, and calibration provide complementary information about prognostic accuracy. What is clear is that investigators will need to look beyond just biomarkers measured at presentation and statistical tests of association—relative risks, hazard ratios, and odds ratios—to make a compelling argument for the clinical utility of an alternative marker in cardiovascular risk prediction. Therefore, we present a modest update to the figure originally proposed by Morrow and Braunwald (1) to account for the findings from Eggers' work and others cited here, on the basis of the increasing recognition of the timing of biomarker measurement after ACS and new statistical methodology (Fig. 1).
View larger version (41K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1 Proposed Updated Biomarker Prognostic Profile in Acute Coronary Syndromes
Darker shades of colors represent greater prognostic information, and lighter shades represent less or equivocal prognostic information at the time point tested. The front star represents measurements during the first days of hospital stay. The back star represents measurements 6 weeks to 6 months after initial presentation. BNP = B-type natriuretic peptide; eGFR = estimated glomerular filtration rate; HbA1c = hemoglobin A1c; hsCRP = high-sensitivity C-reactive protein; NT-proBNP = amino terminal pro–B-type natriuretic peptide. Modified from Morrow and Braunwald (1).
|
|
 |
Footnotes
|
|---|
Dr. deFilippi reports speaking/consulting fees (<$25,000) each for Roche Diagnostics, Siemens, BG Medicine, and Critical Diagnostics and research funding (>$25,000) each from Roche Diagnostics, Siemens, BG Medicine, and Critical Diagnostics.
* Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology. 
|
References
|
|---|
1. Morrow DA, Braunwald E. Future of biomarkers in acute coronary syndromes: moving toward a multimarker strategy Circulation 2003;108:250-252.[Free Full Text]2. Morrow DA, Cannon CP, Jesse RL, et al. National Academy of Clinical Biochemistry Laboratory Medicine practice guidelines: clinical characteristics and utilization of biochemical markers in acute coronary syndromes Circulation 2007;115:e356-e375.[Free Full Text] 3. Jernberg T, Lindahl B, James S, Larsson A, Hansson LO, Wallentin L. Cystatin C: a novel predictor of outcome in suspected or confirmed non–ST-elevation acute coronary syndrome Circulation 2004;110:2342-2348.[Abstract/Free Full Text] 4. Morrow DA, Sabatine MS, Brennan ML, et al. Concurrent evaluation of novel cardiac biomarkers in acute coronary syndrome: myeloperoxidase and soluble CD40 ligand and the risk of recurrent ischaemic events in TACTICS-TIMI 18 Eur Heart J 2008;29:1096-1102.[Abstract/Free Full Text] 5. Sabatine MS, Morrow DA, de Lemos JA, Jarolim P, Braunwald E. Detection of acute changes in circulating troponin in the setting of transient stress test-induced myocardial ischaemia using an ultrasensitive assay: results from TIMI 35 Eur Heart J 2009;30:162-169.[Abstract/Free Full Text] 6. Sabatine MS, Morrow DA, Higgins LJ, et al. Complementary roles for biomarkers of biomechanical strain ST2 and N-terminal prohormone B-type natriuretic peptide in patients with ST-elevation myocardial infarction Circulation 2008;117:1936-1944.[Abstract/Free Full Text] 7. Sabatine MS, Morrow DA, de Lemos JA, et al. Multimarker approach to risk stratification in non-ST elevation acute coronary syndromes: simultaneous assessment of troponin I, C-reactive protein, and B-type natriuretic peptide Circulation 2002;105:1760-1763.[Abstract/Free Full Text] 8. Westerhout CM, Fu Y, Lauer MS, et al. Short- and long-term risk stratification in acute coronary syndromes: the added value of quantitative ST-segment depression and multiple biomarkers J Am Coll Cardiol 2006;48:939-947.[Abstract/Free Full Text] 9. Anderson JL, Adams CD, Antman EM, et al. ACC/AHA 2007 guidelines for the management of patients with unstable angina/non–ST-elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 2002 Guidelines for the Management of Patients With Unstable Angina/Non–ST-Elevation Myocardial Infarction) J Am Coll Cardiol 2007;50:e1-e157.[Free Full Text] 10. Scirica BM, Morrow DA, Cannon CP, et al. Clinical application of C-reactive protein across the spectrum of acute coronary syndromes Clin Chem 2007;53:1800-1807.[Abstract/Free Full Text] 11. Ridker PM, Cannon CP, Morrow D, et al. C-reactive protein levels and outcomes after statin therapy N Engl J Med 2005;352:20-28.[CrossRef][Web of Science][Medline] 12. Lindahl B, Lindback J, Jernberg T, et al. Serial analyses of N-terminal pro-B-type natriuretic peptide in patients with non–ST-segment elevation acute coronary syndromes: a Fragmin and fast Revascularisation during In Stability in Coronary artery disease (FRISC)-II substudy J Am Coll Cardiol 2005;45:533-541.[Abstract/Free Full Text] 13. Zethelius B, Berglund L, Sundstrom J, et al. Use of multiple biomarkers to improve the prediction of death from cardiovascular causes N Engl J Med 2008;358:2107-2116.[CrossRef][Medline] 14. Eggers KM, Lagerqvist B, Venge P, Wallentin L, Lindahl B. Persistent cardiac troponin I elevation in stabilized patients after an episode of acute coronary syndrome predicts long-term mortality Circulation 2007;116:1907-1914.[Abstract/Free Full Text] 15. Morrow DA, de Lemos JA, Blazing MA, et al. Prognostic value of serial B-type natriuretic peptide testing during follow-up of patients with unstable coronary artery disease JAMA 2005;294:2866-2871.[Abstract/Free Full Text] 16. Eggers KM, Lagerqvist B, Venge P, Wallentin L, Lindahl B. Prognostic value of biomarkers during and after non–ST-segment elevation acute coronary syndrome J Am Coll Cardiol 2009;54:357-364.[Abstract/Free Full Text] 17. Pencina MJ, D'Agostino Sr. RB, D'Agostino Jr. RB, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond Stat Med 2008;27:157-172discussion 207–12.[CrossRef][Web of Science][Medline] 18. Cook NR. Use and misuse of the receiver operating characteristic curve in risk prediction Circulation 2007;115:928-935.[Abstract/Free Full Text] 19. Ridker PM, Paynter NP, Rifai N, Gaziano JM, Cook NR. C-reactive protein and parental history improve global cardiovascular risk prediction: the Reynolds Risk Score for men Circulation 2008;118:2243-2251.[Abstract/Free Full Text]
Related Articles
-
Prognostic Value of Biomarkers During and After Non–ST-Segment Elevation Acute Coronary Syndrome
- Kai M. Eggers, Bo Lagerqvist, Per Venge, Lars Wallentin, and Bertil Lindahl
J. Am. Coll. Cardiol. 2009 54: 357-364.
[Abstract]
[Full Text]
[PDF]
-
Inside This Issue
J. Am. Coll. Cardiol. 2009 54: A28.
[Full Text]
[PDF]
|
Top
References